Compare XPUs

Select up to 5 XPUs to compare side-by-side

Select XPUs to Compare

Clear all (4)Jump to results

Filter by Vendor

Showing 72 XPUs • 4 selected

Alibaba

Hanguang 800

AMD

MI100

23.1 TFLOPs

AMD

MI210

181 TFLOPs

AMD

MI250X

383 TFLOPs

AMD

MI300A

980.6 TFLOPs

AMD

MI300X

1,307 TFLOPs

AMD

MI325X

1,400 TFLOPs

AMD

MI350X

2,100 TFLOPs

AMD

Radeon PRO W7900

122 TFLOPs

AMD

Radeon RX 7900 XT

104 TFLOPs

AMD

Radeon RX 7900 XTX

122 TFLOPs

AWS

Inferentia2

190 TFLOPs

AWS

Trainium

190 TFLOPs

AWS

Trainium2

680 TFLOPs

Baidu

Kunlun II

Biren Technology

BR100

Cambricon

MLU370

256 TFLOPs

Cerebras

WSE-3

Enflame Technology

CloudBlazer T20

Etched

Sohu

10,000 TFLOPs

FuriosaAI

Warboy

Google

TPU v4

275 TFLOPs

Google

TPU v5e

197 TFLOPs

Google

TPU v5p

459 TFLOPs

Google

TPU v6e (Trillium)

918 TFLOPs

Graphcore

Bow IPU

Graphcore

IPU-M2000

Groq

LPU Inference Engine

Huawei

Ascend 910B

Iluvatar CoreX

BI-V150

300 TFLOPs

Intel

Data Center GPU Max 1100

177 TFLOPs

Intel

Data Center GPU Max 1550

419 TFLOPs

Intel Habana

Gaudi 2

432 TFLOPs

Intel Habana

Gaudi 3

1,835 TFLOPs

Multi-Metric Comparison

Relative performance across 5 key metrics (normalized to 100 = best in comparison)

Compute Performance (BF16)

Memory Capacity

Power Consumption

Power Efficiency

Specifications

Specification	NVIDIA L4	AMD MI210	SambaNova SN40L	NVIDIA GeForce RTX 4060 Ti
Architecture	Ada Lovelace	CDNA 2	Reconfigurable Dataflow Unit (RDU)	Ada Lovelace
Form Factor	PCIe	PCIe	—	PCIe
VRAM	24 GB	64 GB	640 GB	8 GB
Memory Bandwidth	300 GB/s	1,638 GB/s	—	288 GB/s
TFLOPs (FP32)	30.3	45.3	—	22.1
TFLOPs (FP16)	242	181	—	44.2
TFLOPs	121	181	—	44.2
TFLOPs (FP8)	485	—	—	—
TDP	72 W	300 W	700 W	160 W
Launch Date	Mar 2023	Jan 2022	Jan 2024	May 2023

Efficiency Metrics

Metric	L4	MI210	SN40L	GeForce RTX 4060 Ti
TFLOPs per Watt (FP32-eq)	0.84	0.30	—	0.14
Memory Bandwidth per GB	12.5 GB/s	25.6 GB/s	—	36.0 GB/s

Performance Equivalence

How many units of each GPU are needed to match the performance of the others?

To match 1x NVIDIA L4

AMD MI210

Compute (FP32-eq)

0.67x

MI210 is 1.50x faster

FP32 Compute

0.67x

MI210 is 1.50x faster

VRAM

0.38x

MI210 has 2.67x more

Memory Bandwidth

0.18x

MI210 has 5.46x more

SambaNova SN40L

VRAM

0.04x

SN40L has 26.67x more

NVIDIA GeForce RTX 4060 Ti

Compute (FP32-eq)

2.74x

Need 2.74x GeForce RTX 4060 Ti

FP32 Compute

1.37x

Need 1.37x GeForce RTX 4060 Ti

VRAM

3.00x

Need 3.00x GeForce RTX 4060 Ti

Memory Bandwidth

1.04x

Need 1.04x GeForce RTX 4060 Ti

To match 1x AMD MI210

NVIDIA L4

Compute (FP32-eq)

1.50x

Need 1.50x L4

FP32 Compute

1.50x

Need 1.50x L4

VRAM

2.67x

Need 2.67x L4

Memory Bandwidth

5.46x

Need 5.46x L4

SambaNova SN40L

VRAM

0.10x

SN40L has 10.00x more

NVIDIA GeForce RTX 4060 Ti

Compute (FP32-eq)

4.10x

Need 4.10x GeForce RTX 4060 Ti

FP32 Compute

2.05x

Need 2.05x GeForce RTX 4060 Ti

VRAM

8.00x

Need 8.00x GeForce RTX 4060 Ti

Memory Bandwidth

5.69x

Need 5.69x GeForce RTX 4060 Ti

To match 1x SambaNova SN40L

NVIDIA L4

VRAM

26.67x

Need 26.67x L4

AMD MI210

VRAM

10.00x

Need 10.00x MI210

NVIDIA GeForce RTX 4060 Ti

VRAM

80.00x

Need 80.00x GeForce RTX 4060 Ti

To match 1x NVIDIA GeForce RTX 4060 Ti

NVIDIA L4

Compute (FP32-eq)

0.37x

L4 is 2.74x faster

FP32 Compute

0.73x

L4 is 1.37x faster

VRAM

0.33x

L4 has 3.00x more

Memory Bandwidth

0.96x

L4 has 1.04x more

AMD MI210

Compute (FP32-eq)

0.24x

MI210 is 4.10x faster

FP32 Compute

0.49x

MI210 is 2.05x faster

VRAM

0.13x

MI210 has 8.00x more

Memory Bandwidth

0.18x

MI210 has 5.69x more

SambaNova SN40L

VRAM

0.01x

SN40L has 80.00x more

Pricing

Price Type	L4	MI210	SN40L	GeForce RTX 4060 Ti
CAPEX (Street Price)	$4,000	$6,000	—	—
OPEX (per hour)	$0.80/hr	—	—	—
Price per TFLOPs (FP32-eq)	$66	$66	—	—